Proposed is a longitudinal cross-site comparison study of syringe exchange program (SEP) operational characteristics and changes in individual HIV risk behaviors among injection drug users (IDUs). The overall aim is to determine how differences in SEP operational characteristics may impact changes in HIV risk behavior among SEP clients. Recent reviews of SEP studies suggest that inattention to SEP operational characteristics have biased interpretations of program effectiveness. Yet, few SEP studies provide program information such as syringe distribution protocols and availability of ancillary services. This lack of detail is particularly problematic since SEPs must set policies related to syringe distribution protocol, limits on syringes, and availability of complementary ancillary services without benefit of empirical data on the impact of these choices. To address this problem we propose the following specific aims: (1) To compare changes in HIV risk behaviors among IDUs who use SEPs with different syringe distribution protocols, (2) To compare changes in HIV risk behaviors among IDUs who use a SEP with a limit on the number of syringes that can be exchanged with IDUs who use an SEP with no such limit, and (3), To compare changes in HIV risk behaviors and utilization of medical, social, and substance use services among IDUs who use an SEP that provides complementary ancillary medical services as compared to IDUs who use a SEP without such ancillary services. To accomplish these aims, we will compare SEP client risk behaviors at SEPs with the following operational characteristics: la) one-for-one exchange versus one-for-one plus extra syringes; lb) one-for-one exchange versus need-based distribution, lc) one-for-one plus extra syringes versus need-based distribution; 2) limits on number of syringes exchanged per visit versus no limits on syringes exchanged; and 3) availability of complementary ancillary services (e.g. HIV screening and medical provider, including nurses and/or physician assistant) at SEP sites versus no ancillary services. Each of the 5 selected SEPs is matched to another program on two of the three operational variables of interest, thus permitting comparisons on the third variable. 1,500 IDUs (750 SEP clients; 750 non-SEP using IDUs) will be recruited, interviewed, and followed for 6 months in three waves of data collection beginning in Year 01 and ending in Year 05 from these SEPs and their catchment areas.